8/8/2019 Aircraft Wake Vortex Scenarios

1/5

Engineering NotesENGINEERING NOTES are short manuscripts describing new developments or important results of a preliminary nature. These Notes should not exceed 2500words (where a figure or table counts as 200 words). Following informal review by the Editors, they may be published within a few months of the date of receipt.Style requirements are the same as for regular contributions (see inside back cover).

Aircraft Wake Vortex Scenarios

Simulation Package for Takeoff

and Departure

Frank Holzpfel and Jan Kladetzke

Deutsches Zentrum fr Luft- und Raumfahrt,

82234 Weling, Germany

Swantje Amelsberg

Technische Universitt Berlin, 10587 Berlin, Germany

Helge Lenz and Carsten SchwarzDeutsches Zentrum fr Luft- und Raumfahrt,

38108 Braunschweig, Germany

and

Ivan De Visscher

Universit Catholique de Louvain,

1348 Louvain-la-Neuve, Belgium

DOI: 10.2514/1.39346

I. Introduction

AIRCRAFT-GENERATED wake vortices pose a potential risk

to following aircraft in various flight phases, whereas mostwake-vortex encounters are reported for approach and landing andfor takeoff and climb [1]. The wake-vortex aircraft separationstandards [2] established in the 1970s increasingly degrade aviationefficiency when traffic congestion limits airport capacity duringlanding and takeoff. The most rapid growth scenario within aEUROCONTROL study [3] indicates that in the year 2025, 60European airports could be congested, and as a result, 3.7 millionflights per year could not be met.

Research has shown that the transport and persistence of wakevortices are highly dependent on meteorological conditions [4,5], sothat in many cases the separationstandardsare overconservative. Forsingle-runway operations, analyses [68] suggest that above acertain crosswind threshold, vortices are blown out of the flightcorridor and pose no further threat to following aircraft. The

European Union (EU) project CREDOS (Crosswind-ReducedSeparations for Departure Operations) intends to demonstrate theoperational feasibility of a concept of operations that uses measuresof the prevailing crosswind component to allow temporary

suspension of theneed to apply wake turbulenceseparationsbetweensuccessive departing aircraft.

The focus on the combination of crosswind and departures hassignificant advantages:

1) The follower aircraft is still on the ground when the controllerschedules the separation. So the controller always has the possibilityto extend the separation without requiring the pilot to make amaneuver. This beneficial situation also reduces the time horizon forwhich crosswind conditions must be anticipated.

2) In contrast to arrival situations, the leader aircraft is generallyfaster and so the actual separations tend to increase.

This Note describes the WakeScene-D software package (Wake

Vortex Scenarios Simulation Package for Departure) that has beendeveloped for comprehensive airspace simulations of takeoff anddeparture and the related wake-vortex-induced risks at Frankfurtairport [9]. Within CREDOS, WakeScene-D is used for thefollowing:

1) Support the definition of suitable crosswind criteria that allowreducing aircraft separations.

2) Identify critical parameter combinations.3) Perform risk analyses, taking into account a broad range of

variables that determine the probability and risk of a wake-vortexencounter.

WakeScene-D is an extension of WakeScene that has beendevelopedfor approach andlanding and is described in detail in [10].WakeScene-D estimates the probability to encounter wake vorticesin different traffic and crosswind scenarios using Monte Carlosimulation in a domain ranging from the runway to an altitude of3000 ft above ground. In cases with potential wake encounters, allrelevant parameters can be provided to VESA (vortex encounterseverity assessment) [1012], which may subsequently performdetailed investigations of the severity of the encounter. WakeScene-D consists of elements that model traffic mix, aircraft trajectories,meteorological conditions, wake-vortex evolution, and potentialhazard area. The process and dataflows are controlled and evaluatedby the MATLAB-based environment MOPS (Multi-ObjectiveParameter Synthesis) [13].

In the following a survey on the operating sequence ofWakeScene-D is given and the employed submodels and databasesare described. The determination of crosswind thresholds allowingfor reduced aircraft separations is the subject of ongoing

investigations that will be documented in a future publication.First-application examples of the software package can be found in[9].

II. Survey on Operating Sequence

The flowchart depicted in Fig. 1 sketches the operating sequenceof WakeScene-D. Via simulation control (MOPS), the types of thegenerator aircraft and follower aircraft, the departure routes, and anumber of aircraft and pilot parameters are selected. The trajectorymodel provides time, speed, position, attitude, lift, and mass ofgenerator and follower aircraft along the flight paths. Wake-vortexevolution is predicted within control gates that are released along theflight path of the wake-vortex generator aircraft in predefined timeincrements (e.g., 5 s). Thegates orientations are perpendicular to the

Presented at the 8th AIAA Aviation Technology, Integration, andOperationsConference, Anchorage,AK, 1419 September2008; received 25June 2008; revision received 15 October 2008; accepted for publication 17October 2008. Copyright 2008 by DLR, German Aerospace Center.Published by the American Institute of Aeronautics and Astronautics, Inc.,with permission. Copies of this paper may be made for personal or internaluse, on conditionthat thecopierpay the$10.00 per-copy feeto the CopyrightClearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923; includethe code 0021-8669/09 $10.00 in correspondence with the CCC.

Research Scientist, Institut fr Physik der Atmosphre, Oberpfaffenho-

fen; frank.holzaepfel@dlr.de.Institut fr Robotik und Mechatronik, Oberpfaffenhofen.Institut fr Flugfhrung.Institut fr Flugsystemtechnik.

Data available online athttp://www.eurocontrol.int/eec/credos/(retrievedOctober 2008).

JOURNAL OF AIRCRAFTVol. 46, No. 2, MarchApril 2009

713

http://dx.doi.org/10.2514/1.39346http://www.eurocontrol.int/eec/credos/http://www.eurocontrol.int/eec/credos/http://dx.doi.org/10.2514/1.39346

8/8/2019 Aircraft Wake Vortex Scenarios

2/5

aircraft true heading and perpendicular to the flight-path angle (seeFig. 2). Based on vertical profiles of wind speed and direction, airdensity, virtual potential temperature, turbulent kinetic energy, andeddy dissipation rate (meteorological database) and aircraft position,

speed, attitude,lift, andspan (trajectory model)at onegate, thewake-vortex model simulates the development of wake-vortex trajectories,circulation, vortex-core radius, and attitude of wake-vortex axes. Thesimplified hazard-area prediction model (SHAPE) computes thedistance between wake vortex and follower aircraft within each gateand discriminates between potentially critical cases and cases inwhich safe and undisturbed flight is guaranteed. From all these data,MOPS computes defined criteria such as minimal distance betweenthe wake vortex and follower aircraft and the respective vortexcirculation and height, which are interpolated between the gates andstatistically analyzed. Finally, data needed for further investigationswith VESA are deduced and stored. The results are optionallyvisualized in graphs of the statistics, 2-D and 3-D views (see Fig. 2),or animations of the approaches of subsequent aircraft.

III. Employed Models and Databases

The submodels anddatabasesthat are employed by WakeScene-Dare briefly introduced in the subsequent sections, followed by a briefestimation of the related uncertainties.

A. Meteorological Database

The variety of parameter combinations observed in the planetaryboundary layer and their transformation on wake-vortex behaviorlead to a significant manifold of situations. To capture this diversity,an extensive one-year simulation of realistic meteorologicalconditions has been produced for theFrankfurt terminal areawith thenonhydrostatic mesoscale weather-forecast model system NOWVIV(nowcasting wake-vortex impact variables [4]). NOWVIV

comprises a full physics package, including boundary-layerturbulence; surface energy and momentum balance; soil physics;radiation processes, including cloud effects; cumulus convection;and cloud physics [14].

NOWVIV has previously been successfully employed forpredictions of wake-vortex environmental parameters in five fieldcampaigns [10]. The one-year meteorological database has beenvalidated against a 30-year wind climatology, and a 40-day subsethas been compared with ultrasonic anemometer, SODAR/RASS(sound detection and ranging/radio acoustic sounding system), andlidar (light detection and ranging) measurement data [15].Assessments of wake prediction skill based on predictions ofmeteorological conditions with NOWVIV can be found in [8,16].

The database consists of about 1:3 106 vertical profiles ofmeteorological data at locations separated by 1 n mile and an outputfrequency of 10 min. The meteorological quantities comprise thethree wind components, air density, virtual potential temperature,turbulent kinetic energy, eddy dissipation rate, and pressure.

B. Trajectory Model

The risk of encountering a wake vortex is strongly correlated withthe actual flight paths of the vortex-generating aircraft and theencountering aircraft in space and time. Aircraft trajectories are

modeled beginning on runway 25R along five different standarddeparture routes (TOBAK2F, BIBOS6F, SOBRA1F, ANEKI5F,and DKB2F) until 3000 ft above ground. The departure routesdiverge at four waypoints leading from northwesterly up tosoutheasterly flight directions. Figure 2 shows exemplary aircrafttrajectories for a southeasterly (leader) and a westerly (follower)route.

A large number of environmental- andaircraft-specific parametersinfluence an aircraft trajectory. The trajectory model (see [17] fordetails) simulates the impact of the most relevant parameters, whichare the selected runway and the standard departure route;meteorological conditions, including air temperature, density,pressure, wind direction, and strength; aircraft type; aircraft takeoffweight; takeoff thrust that can be either takeoff go-around thrust orflex takeoff thrust (reduced thrust); start position on the runway; andpilot behavior that is described as a control model consideringdelayed pilot reaction time and a cross-track error.

These factors are varied within defined boundaries and givenprobability distributions employing Monte Carlo simulation togenerate a set of trajectories for different aircraft types and departureconditions. The aircraft trajectory is adequately described with theequations of motion for three translational degrees of freedom.

A deterministic verification has been accomplished by comparingresults of the trajectory model with high-fidelity simulation data ofdepartures that were simulated on the certified A330-300 full-flightsimulator (A330-FFS) at Technische Universitt Berlin. Further-more, a statistical validation was performed by comparingMonte Carlo simulation results of the trajectory model with 20,000measured departures at Frankfurt airport provided by DFS Deutsche

Flugsicherung, GmbH.Figure 3 shows exemplary results of 1000 simulated A320

departures. The simulations are based on variations of pilot behavior,aircraft weight, thrust mode, and wind conditions. In Fig. 3 theresulting mean trajectory and its standard deviations (black) arecompared with the respective measurement data (gray). Theagreement of the lateral flight path, the climb profile, and the speedprofile is sound.

C. Wake-Vortex Prediction Models

WakeScene provides a choice between two different parametricwake-vortex prediction models. These are the deterministic two-phase wake-vortex decay model (D2P) and the deterministic wake-vortex model (DVM). Both vortex models have been validated fordeparting aircraft by evaluating statistics of the deviations betweenmeasured and predicted wake-vortex behavior employing dataacquired during the two CREDOS field measurement campaigns at

Fig. 1 WakeScene-D flowchart. Arrows denote the data flow.

714 J. AIRCRAFT, VOL. 46, NO. 2: ENGINEERING NOTES

8/8/2019 Aircraft Wake Vortex Scenarios

3/5

Frankfurt airport. The alternative application of two wake-vortexmodels allows assessing the sensitivity of WakeScene-D results onwake-vortex parameterizations.

A few adaptations were necessary to comply with the architecturedevised for WakeScene-D. The control gates in which the vorticesevolve (seeFig. 2) areinclined by theflight-path anglewith respectto the vertical direction. For curved flight the vortices are initializedin positions rotated by the bank angle such that the vorticesdescend in a direction tilted by . The wake-vortex transport byheadwind or tailwind is modeled by the respective transport of thegates. Because the gates have arbitrary orientations and movethrough the space, the determination of the closest distance between

wake vortex and follower aircraft requires somewhat complexcalculations.

The probabilistic vortex model P2P, which constitutes the basis ofits deterministic version D2P, is described in detail in [18].Applications, assessments, and further developments are reported in[16,19,20]. In total, P2P has been validated against data of over 1400cases gathered in two United States and five European measurementcampaigns. D2P accounts for the effects of wind, axial, andcrosswind shear;turbulence; stable thermal stratification; and groundproximity. Figure 4 delineates a comparison between measuredwake-vortex positions and circulation and the predictions of the twowake-vortex models. Note the effective lateral transport of the

Fig. 2 Perspective view of trajectories of wake-generating aircraft (blue) and follower aircraft (magenta) together with wake-vortex positions(starboardvortex green,port vortexred). Projections of aircraft trajectories on vertical and horizontalplanes and a numberof gatesused for wake-vortexprediction are displayed for convenience.

0 1000 2000 3000 4000 5000 6000 7000200

0

200

400

600

800

x [m]

y[m]

Mean DFSData A320

Std+/ DFSData

Mean Traj. Model A320

Std+/ Traj. Model

0 1000 2000 3000 4000 5000 6000 70000

200

400

600

800

1000

Height[m]

x [m]

0 1000 2000 3000 4000 5000 6000 70000

20

40

60

80

100

120

TrueAirspeed[m/s]

x [m]

Fig. 3 Statistics of 1000 departures of A320 aircraft. Lateral and vertical positions as well as speed profile.

J. AIRCRAFT, VOL. 46, NO. 2: ENGINEERING NOTES 715

8/8/2019 Aircraft Wake Vortex Scenarios

4/5

vortices in a case with a crosswind of about 4 m=s. Further outputprovided to VESA includes vortex-core radii and the interceptionangles between aircraft flight path and vortex axis: the so-called

encounter angles.The DVM is the new wake-vortex predictor software developed

by the Universit Catholique de Louvain, establishing a step forwardin terms of robustness, modularity, and performance. It is based onthe numerical methodology and the physical models of the VortexForecast System, originally developed by an international team [21]and further improved and calibrated (against two United States andtwo EU campaigns and against large eddy simulation) since 2002 inthe framework of EU-funded projects (I-Wake, ATC-Wake [22],AWIATOR, WakeNet2-Europe [1], FLYSAFE, FAR-Wake, andCREDOS). The DVM accounts for the effects of wind transport(cross and axial), wind shear, decay due to turbulence, stratification,and ground proximity [23]. It also includes improvements regardingthe evaluation of the vertical profiles of environmental conditionsand of the in-ground-effect model. The probabilistic wake-vortex

model is using the DVM in a Monte Carlo approach, taking intoaccount the uncertainties and variations of the impact parametersfrom theaircraft andmeteorological side andof some physical modelcoefficients of the DVM.

D. Hazards-Area Module

The hazards-area module computes the distance of the followeraircraft (centerof gravity) to thevortex centers withineach gate alongtheflight path. Then theclosest distance between thefolloweraircraftand the vortex pair over all gates along the flight path is determinedby interpolating aircraft trajectories and wake trajectories betweenthe gates. This closest approach is used for further statistical analysiswith WakeScene-D (e.g., vortex age and circulation for this point intime).

To estimate theseverity of thepotential wake-vortex encounter,anarea of interest can optionally be defined around the vortices. If thetrailing aircraft penetrates this area of interest, the wake-vortexencounter is classified as potentially hazardous. This is consideredasa preliminary severity assessment. A corresponding concept calledsimplified hazard areas has been developed in [24] and adapted fortakeoffand departure in [9]. Cases violatingthe areaof interest canbesubject to more detailed severity assessment, which is outside of thescope of WakeScene-D.

E. Estimation of Uncertainty

Any software that may be employed to assess the safety of a wake-vortex advisory system must constitute a sufficiently accuraterepresentation of the projected operation. However, for complexrisk-assessment tools, straightforward validation appears to beinfeasible, because the significant manifold of modeled parameterscannot be measured simultaneously and reconstructed consistently

in a simulation. For WakeScene-D the identification of the relevantprocesses and the definition of the appropriate degree of details withwhich they have to be modeled rely on thorough discussion and

expert opinion. For the validation of the employed submodels, werefer to the studies cited in the previous sections [8,10,1517,19,20].

Here, we exemplarily perform a simple estimation of uncertaintyfor the most important parameter, which is lateralvortex transport, ata vortex generator height of about 50 m, which is within the mostcritical height range of 100 m above ground [1,9]. For B744 andA343 aircraft, the difference of predicted and measured standarddeviations of lateral aircraft position amounts to ac 6:8 m. Forthese aircraft types, the median RMS deviation of measured andpredicted wake-vortex positions employing the D2P model has beenestimated to vort 19:2 m [20]. For our purposes here, the statisticsonly need to be accurate on average; that is, the predictions do notneed to be correct in space and time. Because the vortex positionuncertainty is dominated by spatial and temporal variations betweenthe measured crosswind and the crosswind that is sensed by the

vortices, the latter estimation is outmost conservative. Therefore, itcan be understood that the vortex prediction uncertainty also impliesuncertainties from numerical weather prediction. Assuming thatthese uncertainties are independent and statistically stationary, theoverall uncertainty amounts to

tot 2ac

2vort

p 20:3 m

For this scenario and 120-s-old vortices, WakeScene-D determines avortex spreading with obs 337:3 m. Hence, a very conservativeestimation of the relative uncertainty of lateral vortex positions inground proximity can be estimated to 6.0%.

IV. Monte Carlo Simulation

Finally, we briefly describe the setup of a Monte Carlo simulationwith a typical sample size of 500,000 aircraft pairings. A306, A310,A333, A343, B744, and B772 aircraft are used as vortex generatorsand A320, AT45, B733, and CRJ are used as followers. The trafficmixis modeled according to thestatistics of Frankfurt airport in 2006[25]. The following parameters of the generator and the followeraircraft are randomly distributed: start point, takeoff weight, thrustmode, departure-route combination, trajectory deviation, and pilot-delay parameter.

In the reference case, the follower aircraft obey the 120 sInternational Civil Aviation Organization separation and meteoro-logical data of the full one-year database are employed. The baselinecase may then be compared with situations with reduced temporalaircraft separations combined with different crosswind criteria. Thedetermination of appropriate crosswind thresholds allowing forreduced aircraft separations is the subject of ongoing work that will

0

50

100

150

200

250

300

350

0 20 40 60 80 100120140160

z

[m]

t [s]

lidar portlidar starb.

DVMD2P

0

100

200

300

400

500

600

0 20 40 60 80 100120140160

y

[m]

t [s]

0

100

200

300

400

500

600

0 20 40 60 80 10012014016

[m2/s]

t [s]

Fig. 4 Example for evolution of vertical and lateral positions and circulation in a case with a crosswind of about 4 m=s from CREDOS EDDF-1campaign. Measurements by lidar (symbols) and predictions with D2P and DVM wake-vortex models (lines).

716 J. AIRCRAFT, VOL. 46, NO. 2: ENGINEERING NOTES

8/8/2019 Aircraft Wake Vortex Scenarios

5/5

be documented in a futurepublication. Thefirst applicationexamplesof the software package can be found in [9].

V. Conclusions

WakeScene-D, a software package to determine wake vortexencounter probabilities for departures is described. The componentsof WakeScene-D that model traffic mix, aircraft trajectories,meteorological conditions, wake vortex evolution, and potentialhazard area are specified together with the validation workaccomplished thus far. Further, a brief estimation of relateduncertainties is provided. WakeScene-D is an extension ofWakeScene [10] that has been developed for approach and landing.Compared with this precursor version, major improvements havebeen achieved regarding realistic aircraft trajectory modeling ondifferent standard departure routes. The developed trajectory modelfully considersthe impact of themost relevant parameters for6 heavyvortex generator and 4 medium follower aircraft types. Thecomparison of simulated departures with full-flight simulator dataandwith 20,000measureddeparturesat Frankfurt airport yieldsgoodagreement. Another step ahead has been achieved by modeling thewake vortex behavior within inclined and moving gates. The gatesthat are oriented perpendicularly to the aircraft heading and flight-path angles are released in predefined time steps along the flight path

and are advected through space by the prevailing headwindsrespective tailwinds. The real target of WakeScene-D, thedetermination of appropriate crosswind thresholds allowing forreduced aircraft separations, is the subject of ongoing work and willbe documented in a futurepublication. Thefirst applicationexamplesof the software package can be found in [9].

Acknowledgments

The financial support from the European Union project CREDOS(AST5-CT-2006-030837) and from the DLR, German AerospaceCenter project Wirbelschleppe is greatly acknowledged.

References

[1] Anon., Wake Vortex Research Needs for Improved Wake VortexSeparation Ruling and Reduced Wake Signatures, WakeNet2-Europe,National Aerospace Lab., Rept. NLR-CR-2006-171, Amsterdam,Mar. 2006.

[2] Procedures for Air Navigation ServicesAir Traffic Management,International Civil Aviation Organization, Rept. 4444, ATM/501,Montreal, 2001.

[3] Anon., Challenges to Growth 2004 Report, EUROCONTROL,Rept. CTG04, Brussels, Dec. 2004.

[4] Gerz, T., Holzpfel, F., Bryant, W., Kpp, F., Frech, M., Tafferner, A.,and Winckelmans, G., Research Towards a Wake-Vortex AdvisorySystem for Optimal Aircraft Spacing, Comptes Rendus Physique,Vol. 6, Nos. 45, 2005, pp. 501523.doi:10.1016/j.crhy.2005.06.002

[5] Hallock, J. N., Greene, G. C., and Burnham, D. C., Wake VortexResearchA Retrospective Look, Air Traffic Control Quarterly,Vol. 6, No. 3, 1998, pp. 161178.

[6] de Bruin, A. C., Speijker, L. J. P., Moet, H., Krag, B., Luckner, R., andMason, S., S-WakeAssessment of Wake Vortex Safety, NationalAerospace Lab., Rept. NLR-TP-2003-243, Amsterdam, 2003.

[7] Frech,M.,and Zinner, T.,Concept of WakeVortex Behavior Classes,Journal of Aircraft, Vol. 41, No. 3, 2004, pp. 564570.doi:10.2514/1.55

[8] Frech, M., and Holzpfel, F., Skill of an Aircraft Wake-Vortex ModelUsing Weather Prediction and Observation, Journal of Aircraft,Vol. 45, No. 2, 2008, pp. 461470.doi:10.2514/1.28983

[9] Holzpfel, F., Kladetzke, Amelsberg, S., Lenz, H., Schwarz, C., and

De Visscher, I., Aircraft Wake Vortex Scenarios Simulation for Take-Off and Departure, AIAA Paper 2008-8921, Sept. 2008.

[10] Holzpfel,F., Frech, M., Gerz,T., Tafferner,A., Hahn,K.-U.,Schwarz,C., Joos, H.-D., Korn, B., Lenz, H., Luckner, R., and Hhne, G.,Aircraft Wake Vortex Scenarios Simulation PackageWakeScene,AerospaceScience and Technology,Vol.13,No.1,2009,pp.111.doi:10.1016/j.ast.2007.09.008.

[11] Hhne, G., Luckner, R., and Fuhrmann, M., Critical Wake VortexEncounter Scenarios, Aerospace Science and Technology, Vol. 8,No. 8, 2004, pp. 689701.

doi:10.1016/j.ast.2004.07.005[12] Luckner, R., Hhne, G., and Fuhrmann, M., Hazard Criteria for WakeVortex Encounters During Approach, Aerospace Science andTechnology, Vol. 8, No. 8, 2004, pp. 673687.doi:10.1016/j.ast.2004.06.008

[13] Joos, H.-D., Bals, J., Looye, G., Schnepper, K., and Varga, A., AMultiobjective Optimisation-Based Software Environment for ControlSystems Design, 2002 IEEE International Symposium on Computer Aided Control Systems Design, Inst. of Electrical and ElectronicsEngineers, Piscataway, NJ, 2002, pp. 714.

[14] Grell,G. A.,Emeis, S.,Stockwell,W. R.,Schoenemeyer,T., Forkel, T.,Michalakes, J.,Knoche, R.,and Seidl, W.,Application of a Multiscale,Coupled MM5/Chemistry Model to the Complex Terrain of theVOTALP Valley Campaign, Atmospheric Environment, Vol. 34,No. 9, 2000, pp. 14351453.doi:10.1016/S1352-2310(99)00402-1

[15] Frech,M., Holzpfel, F.,Tafferner, A.,and Gerz, T., High-ResolutionWeather Data Base for the Terminal Area of Frankfurt Airport, Journal of Applied Meteorology and Climatology, Vol. 46, No. 11,2007, pp. 19131932.

[16] Holzpfel, F., and Robins, R. E., Probabilistic Two-Phase AircraftWake-Vortex Model: Application and Assessment, Journal ofAircraft, Vol. 41, No. 5, 2004, pp. 11171126.doi:10.2514/1.2280

[17] Amelsberg, S.,and Luckner, R., Parametric Aircraft TrajectoryModelfor Takeoff and Departure, 1st CEAS European Air and SpaceConference, Berlin, Confederation of European Aerospace SocietiesPaper CEAS-2007-273, Sept. 2007, pp. 785795.

[18] Holzpfel, F., Probabilistic Two-Phase Wake Vortex Decay andTransport Model, Journal of Aircraft, Vol. 40, No. 2, 2003, pp. 323331.doi:10.2514/2.3096

[19] Holzpfel, F., Probabilistic Two-Phase Aircraft Wake-Vortex Model:

Further Development and Assessment, Journal of Aircraft, Vol. 43,No. 3, 2006, pp. 700708.doi:10.2514/1.16798

[20] Holzpfel, F., and Steen, M., Aircraft Wake-Vortex Evolution inGround Proximity: Analysis and Parameterization, AIAA JournalVol. 45, No. 1, 2007, pp. 218227.doi:10.2514/1.23917

[21] Jackson, W.,Yaras, M.,Harvey,J., Winckelmans,G., Fournier,G., andBelotserkovsky, A., Wake Vortex Prediction: An Overview,Transport Canada, Transportation Development Centre, Rept. TP13629E, Montral, Mar. 2001, http://www.tc.gc.ca/innovation/tdc/summary/13600/13629e.htm[retrieved Jan. 2009].

[22] Speijker, L. J. P., Vidal, A., Barbaresco, F., Frech, M., Barny, H., andWinckelmans, G., ATC-Wake: Integrated Wake Vortex Safety andCapacity System,Journal of AirTraffic Control, Vol.49, No. 1,2007,pp. 1732.

[23] Winckelmans, G., Duquesne, T., Treve, V., Desenfans, O., andBricteux, L., Summary Description of the Models Used in the VortexForecast System (VFS), Univ. Catholique de Louvain, Louvain-la-Neuve, Belgium, Apr. 2005.

[24] Schwarz, C. W., and Hahn, K.-U., Full-Flight Simulator Study forWake Vortex Hazard Area Investigation, Aerospace Science andTechnology, Vol. 10, No. 2, 2006, pp. 136143.doi:10.1016/j.ast.2005.09.005

[25] Anon., Frankfurt Airport Luftverkehrsstatistik 2006, Fraport AG,Frankfurt am Main, Germany, 2007.

J. AIRCRAFT, VOL. 46, NO. 2: ENGINEERING NOTES 717

http://dx.doi.org/10.1016/j.crhy.2005.06.002http://dx.doi.org/10.2514/1.55http://dx.doi.org/10.2514/1.28983http://dx.doi.org/10.1016/j.ast.2007.09.008http://dx.doi.org/10.1016/j.ast.2004.07.005http://dx.doi.org/10.1016/j.ast.2004.06.008http://dx.doi.org/10.1016/S1352-2310(99)00402-1http://dx.doi.org/10.2514/1.2280http://dx.doi.org/10.2514/2.3096http://dx.doi.org/10.2514/1.16798http://dx.doi.org/10.2514/1.23917http://www.tc.gc.ca/innovation/tdc/summary/13600/13629e.htmhttp://www.tc.gc.ca/innovation/tdc/summary/13600/13629e.htmhttp://dx.doi.org/10.1016/j.ast.2005.09.005http://dx.doi.org/10.1016/j.ast.2005.09.005http://www.tc.gc.ca/innovation/tdc/summary/13600/13629e.htmhttp://www.tc.gc.ca/innovation/tdc/summary/13600/13629e.htmhttp://dx.doi.org/10.2514/1.23917http://dx.doi.org/10.2514/1.16798http://dx.doi.org/10.2514/2.3096http://dx.doi.org/10.2514/1.2280http://dx.doi.org/10.1016/S1352-2310(99)00402-1http://dx.doi.org/10.1016/j.ast.2004.06.008http://dx.doi.org/10.1016/j.ast.2004.07.005http://dx.doi.org/10.1016/j.ast.2007.09.008http://dx.doi.org/10.2514/1.28983http://dx.doi.org/10.2514/1.55http://dx.doi.org/10.1016/j.crhy.2005.06.002